本文利用一种快速气相色谱仪（zNoseTM电子鼻）对芒果在采后熟化过程中释放的挥发性物质进行了定量的检测及分析，通过提取不同的挥发性物质峰值来对芒果挥发物成分进行标定，并建立数学模型对芒果表皮黑斑率和成熟度进行了检测。文中通过对比分析气味、表皮黑斑率、成熟度、可溶性固形物含量以及呼吸速率等实验数据，发现利用电子鼻检测数据中的峰4和峰5值能有效地对腐烂程度做出判定，峰7值跟成熟度具有很高的相关性，而可溶性固形物指标在芒果成熟度判定中贡献度很小。通过在峰4峰5值与腐烂度之间建立高斯模型，并利用联合阈值对腐烂进行检测判定，其检测准确率超过90%；在峰7值与成熟度之间建立分段指数模型，可有效地对成熟度做出估计，其估计均方根误差可控制在7%以内。

Mango volatiles released during postharvest ripening were quantitatively measured and analyzed with an ultrafast gas chromatography system (zNoseTM). The mango volatiles were characterized by extracting the mass spectral peaks of different volatiles and mathematical models were established to measure the blackspot rate of mango skin and degree of ripeness. Odor, skin blackspot rate, degree of ripeness, soluble solid content, respiration rate, and other experimental data were compared and analyzed. The results showed that peaks 4 and 5 from the electronic nose data could effectively determine the degree of rot, peak 7 exhibited a high correlation with the degree of ripeness, and the soluble solid content showed a small contribution to the determination of the degree of ripeness. Gaussian models were constructed between the degree of rot and values of peaks 4 and 5. The rot of mango was determined using established models and the threshold value and accuracy rate of determination was over 90%. A piecewise exponential model was constructed between the degree of ripeness and value of peak 7, an effective estimation on the degree of ripeness was achieved. The root mean square error of the estimation was controlled within 7%.

国家自然科学基金资助项目(51508229)；江苏省产学研前瞻性联合研究项目(BY2014023-32)；江苏省食品先进制造装备重点实验室开放课题(FM-201406)